Project title: PFAS exposure of humans, animals and environmental compartments: an evidence review map and bibliometric analysis
In this document I provide the results of my Evidence Review Map (Chapter 1 of my thesis). Results are organized in 3 objectives.
What evidence on PFAS has been systematically synthesized?
I aim to reveal what areas in the realm of PFAS health-related and environmental research have been synthesized the most and where syntheses of evidence are lacking.
How robust are systematic syntheses of PFAS evidence?
I examine the included syntheses for their reporting quality and potential biases, in order to assess reliability of reviews’ conclusions, reveal syntheses that should be re-done, and highlight the aspects of review methodology that need to be improved.
How is synthesized PFAS evidence connected?
I examine which countries and institutions are mostly involved in secondary PFAS research and what do the networks between these institutions look like.
PRISMA flowchart
However, we have 42 more SRs found during an update search up to end 2022.
Mapping: What evidence on PFAS has been systematically synthesized?
Humans are the most reviewed subject in SRs on PFAS exposure, with 82 SRs, followed by multiple subjects (12), environmental compartments (8), and animals (7).
“Humans” and “toxicity” were the most frequently used MeSH terms across all subjects, followed by “fluorocarbons” and “chemically induced”.
The number of SRs published annually has steadily increased over time, with a significant spike in 2021.
Recent syntheses are less likely to include quantitative synthesis through meta-analysis.
The increase in published SRs over the years is due to SRs on humans.
Reviews after 2019 tend to focus on PFAS and not on POPs in general.
PFCA and PFSA are still the two most reviewed groups.
PFAS with a short or long carbon chain are less reviewed than PFAS with a medium-length carbon chain.
Species plots:
Critical appraisal: How robust are systematic syntheses of PFAS evidence?
Most SRs included a statement regarding conflicts of interest and few acknowledged the potential presence of COI.
Most SRs reported receiving funding, with the majority funded by non-profit organizations.
Most SRs did not make their raw data or analysis code available, limiting the ability to verify findings or replicate analysis.
Over 50% of the SRs received poor scores (i.e., 0 or 0.5 scores) on a majority of AMSTAR 2 questions, indicating a need for improvement in the rigor and quality of the literature.
The highest scoring assessment questions related to clarity of explanation and reporting of conflict of interest.
The lowest scoring questions related to transparency measures, such as reporting sources of funding of the included studies and list excluded studies.
Many SRs lacked clear research questions and inclusion criteria, established review methods prior to the conduct of the review, and used a comprehensive literature search strategy.
Many SRs also had limited reviewer involvement in study selection and data extraction, and inadequate techniques for assessing the risk of bias and discussing heterogeneity and publication bias.
Only half of the recent SRs follow reporting guidelines.
Recent SRs are more likely to include a conflict-of-interest statement and a funding statement than earlier reviews.
Almost all the reviews are founded by no-profit organizations.
Quality of reviews (AMSTAR2 score) differs according to field of research, publication year, review type, publication source. work in progress
Bibliometric analysis: How are PFAS reviews connected?
work in progress
Figures in the main paper:
Chord diagram of the collaborations between affiliate countries. Example: check pdf file
Chord diagram showing connectedness of review subjects (humans, animals, environment)
Plus more analyses in supplementary materials